Skip to main content
. 2021 Jan 28;12:672. doi: 10.1038/s41467-020-20756-2

Fig. 4. Output of AutoCD for three-strain stable steady state objective.

Fig. 4

a Dendrogram is generated by hierarchical clustering of the adjacency matrices for each model in the three-strain model space. We set the limit of number of levels to 5, in order to show high level groups. Each column of the heatmap represents a possible interaction between state variables, where green indicates the interaction exists for the model and black indicates absence of the interaction. The posterior probability plot shows the average posterior probability within each group of models. b Shows the models with highest posterior probability when subsetted for number of parts expressed, in order of increasing complexity (3, 4, 5 and 6 expressed parts). The bar chart shows the mean model posterior probability across three experiments, represented by the scatter points, the error bars indicate the standard deviation. n = 3, where each replicate consists of 825,000 accepted particles. Bayes factors (BF) are shown for pairwise comparison of the three models and error bars show the standard deviation between three repeat experiments. Model schematics show the interactions between strains (blue, green and red), bacteriocin (red) and QS molecules (purple). Models with two parts showed posterior probability 0.0 and are not shown.